Lepidium persicum, Boissier, 1842

German, Dmitry A., 2014, Taxonomic remarks on some Asian Lepidium s. l. (Brassicaceae), Phytotaxa 186 (2), pp. 97-105 : 99-101

2. Lepidium persicum s. l.

In an excellent treatment of Lepidium for the Flora Iranica, Hedge (1968) recognized four pretty well separated subspecies of L. persicum . Two of them are worthy to be mentioned here.

Lepidium ferganense Korshinsky (1898: 417) . Type indication: “Hab. in decliviis siccis montium argillosis vel lapidosis, alt. 2–4000’. Fergana: prope Dshelabad, 2 (14) Aug. [18]95 fl. fr.; prope Kargisch-mazar ad fl. Karasu, 15 (27) Aug. [18]95 fr.”. Lectotype (designated here):— KYRGYZSTAN. [Jalal-Abad] Dzhelabad, dry loess hills. 2 VIII 95. S. Korshinsky 361 (LE!, isolectotypes (2×) and syntype LE!).

= L. persicum subsp. arianum Hedge (1968: 70 , 71, 343), syn. nov. Type:— AFGHANISTAN. O-Afghanistan: bei Kabul. Tobfels am Scher Darwasah , 1840 m. 7 VI 1950. A. Gilli 1152 (holotype W!).

= L. seravschanicum Ovcz. & Junussov in Junussov (1978: 628), syn. nov. Type :— TAJIKISTAN. Northern slope of Zeravshan range, Pastiv gorge, near Falmaut, stony and gravelly slopes, 1600 m. 16 VIII 1959. S.I. Junussov [Yunusov] 875 (holotype TAD).

Lepidium persicum subsp. arianum was distinguished from the type subspecies, in particular, by longer (4–6 vs. 1–4 mm), divaricate-ascending (vs. usually erect) pedicels and oblong-lanceolate (vs. predominantly linear-oblong), relatively soft stem leaves and was also characterized by more eastern distribution (central to north Afghanistan). It was noted that “… it is certainly nearer to some of the Soviet Central Asian taxa than to subsp. persicum ” ( Hedge, 1968: 71). This statement is definitely true with respect to L. ferganense , highly variable species widely distributed in neighboring Middle Asian Mountains (Tian Shan and Pamir Alai). Hedge’s conclusion can be compared with the note of Botschantzev & Vvedensky (1955: 199) who emphasized morphological variability of L. ferganense and stressed that “… it is very close to Iranian L. persicum Boiss. from which it differs mainly in longer and ascending pedicels”. Comparison of abundant material of both L. persicum subsp. arianum and L. ferganense revealed that morphologically the prior falls within the range of variability of the latter. For this reason, these taxa are synonymized here under the name L. ferganense as long as the species status better reflects distinctness of this entity from L. persicum .

Although Afghanistan was mentioned in the general distribution of L. ferganense in several floras ( Botschantzev & Vvedensky 1955, Vassiljeva 1961, Junussov 1978, Zhou et al. 2001), no localities were ever given and the specimens on which those records were based are unknown to me. With the present synonymization of L. persicum subsp. arianum , occurrence of the species in Afghanistan is confirmed and its distribution in this country is well characterized by the data of Flora Iranica ( Hedge 1968).

Lepidium seravschanicum is also assigned here to synonymy of L. ferganense for the following reasons. It was said to differ from L. ferganense in pubescent (non glabrous or nearly so), oblong-spatulate or narrowly obovate (vs. oblong lanceolate), dentate only apically (not throughout the margin) lower leaves, pubescent (vs. glabrous) sepals, silicles 2–2.5 × 1.5–2.5 (not 2.5–3.5 × 2.3–3) mm, and habitually being more elegant due to longer and more lax inflorescences and having numerous vs. few stems. In fact, all these characters in different combinations appear throughout the distribution range of L. ferganense . For example, relatively small fruits are common in the eastern part of the species range including its locus classicus in the foothills of Fergana range where plants with generally more compact inflorescences occur while in the rest of the distribution area, plants with strongly elongated inflorescences and bigger fruits dominate. Specimens with pubescent leaves are especially common in the western part of the species distribution and plants with less dentate to entire and in general narrower leaves prevail in eastern and northern fragments of the area. Amount of stems varies greatly and can be rather high (up to 12 in some specimens from Kuraminsky range) and pubescent sepals are revealed throughout the range of the species. Thus, L. seravschanicum apparently represents a narrowly distributed (endemic to Seravshan range) race which exhibits a certain fraction of morphological variation of L. ferganense . Taxonomically, it might be treated as a variety of the latter species but specific rank does not seem appropriate for this entity. This agrees with the conclusion of Botschantzev & Vvedensky (1955) who emphasized “the tendency of forming local races” but refrained from splitting L. ferganense into several species, and this approach looks adequate even with much more material collected since then.

In light of this variability, one more deviating form of L. ferganense s. l. should be commented. This is the above mentioned entity distributed in South Uzbekistan and South East Turkmenistan (western Pamir Alai Mts.). It was reported by Vassilczenko (1948: 252) under the invalidly published name “ L. kuhitangi Vass. ” and later by Vinogradova (1974) as L. lacerum (p. p., quoad pl. e Tian Shan & Pamir Alai). Relevant specimens are stored in LE and TASH under these names and partly as L. ferganense . They are peculiar for having dissected, with one to few pairs of lateral lobes, lower and sometimes even middle leaves which makes them superficially resembling L. lacerum and, due to usually well developed indumentum of both leaf surfaces, also to L. subcordatum Botschantzev & Vvedensky (1941: 12) . However, as noted above, the whole set of characters prevents identifying them as L. lacerum and indicates closest affinity to L. ferganense . The same is true with regards of L. subcordatum which has ovate, cordate at base silicles, more densely pubescent leaves, smaller dimensions of the whole plants, etc. Relationship of the discussed specimens to L. ferganense is further confirmed by the continuous gradient in degree of leaf dissection from entire and shallowly crenate-toothed to deeply incised and finally lyrate-pinnatifid. Individuals with at least deeply incised leaves occur in other parts (e. g., Turkestanian range) of distribution area of this species. It can be suggested that such deviating morphology of specimens from western Pamir Alai might be a reflection of possible former introgression with L. lacerum or L. subcordatum or that “ L. kuhitangi ” is a race not yet completely diverged from L. ferganense and deserving recognition as its subspecies but these hypotheses should be checked by molecular study of the whole group. Until it is done it seems wiser considering relevant specimens as belonging to L. ferganense without giving them any taxonomic status.

Another subspecies of L. persicum established in the Flora Iranica to be mentioned here is subsp. macrocarpum Hedge (1968: 70 , 71, 344).

Lepidium macrocarpum ( Hedge) D.A. German , comb. et stat. nov. ≡ L. persicum subsp. macrocarpum Hedge (1968: 70 , 71, 344). Lectotype (designated here):— AFGHANISTAN. Prov. Mazar-i-Shariff: south east of Tashqurghan on road to Samangan (Aybak), rocky slopes of gorge. Alt. c. 600 m. 9 th June 1962. Ian Hedge & Per Wendelbo No. W. 3960 (E: http://data.rbge.org.uk/herb/E00320596!; isolectotypes BG, LE!).

As the type was indicated by reference to the two parts of relevant gathering (in BG and E), these specimens are syntypes ( ICN Art. 40.1, Note 1; McNeill et al. 2012) one of which is designated here as lectotype .

This taxon was separated from other subspecies of L. persicum by lack of indumentum, longer and entire leaves, and, most important, bigger, 3–4 × 2–3 mm, fruits. It should be added that it differs from the closest species, L. persicum and L. ferganense , in pedicels pubescent strictly adaxially (vs. glabrous or variously pubescent), elliptic (vs. broadly ovate, broadly elliptic or round), much less carinate silicles with lanceolate to linear-lanceolate (vs. broadly lanceolate to elliptic) replum, and at least distally margined (vs. not margined) seeds. Taking into consideration the paucity of features separating the taxa in this group, such a set of characters is definitely enough to justify raising its taxonomic rank to a species.

Lepidium macrocarpum is known based on few gatherings exclusively from the type locality south-east of Kholm (Tashqorgan) in North Afghanistan ( Hedge 1968) and thus is considered a very local endemic. In fact, its distribution area seems to be slightly wider because two collections of “ L. ferganense ” from adjacent Tajikistan (both from Khodzha-Kazyan Mts. in ca. 80 km NNE from locus classicus of L. macrocarpum ) morphologically corresponds the latter species rather than L. ferganense s. l. Ecologically, L. macrocarpum is confined to limestone rocks, as evidenced by the label data of specimens from both Afghanistan and Tajikistan.

Tajikistanian specimens examined:— TAJIKISTAN. Tajik SSR. Khodzha-Kazyan range, southern slope of mt. Taka-Kamar , pistachio communities on limestone slopes, 1000 m, 7 October 1976, G. Kinzykayeva & T. Kochkareva s.n. (exsicc. Fl. URSS No. 6672. Lepidium ferganense Korsh. ) ( BP!, BRNU!, G!, KW!, M!, MHA!, MW!, NS!, PR 709136!, W 1995-05895!); South Tajikistan. Khodzha-Kazyan Mts. to the east of Shaartuz. Crevices of limestone rocks. 23 May 1960. V.P. Botschantzev & T.V. Egorova 335 ( LE!) .